From mboxrd@z Thu Jan 1 00:00:00 1970 From: seb Subject: mount freeze, acl system Date: Mon, 22 Jul 2002 16:41:22 +0200 Sender: nfs-admin@lists.sourceforge.net Message-ID: <3D3C1992.7090403@h3d.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Return-path: Received: from elvis.h3d.com ([195.167.192.132] helo=mail.h3d.com) by usw-sf-list1.sourceforge.net with esmtp (Exim 3.31-VA-mm2 #1 (Debian)) id 17WeNu-0004M7-00 for ; Mon, 22 Jul 2002 07:42:02 -0700 Received: from smtp3.h3d.com (ATuileries-101-2-1-75.abo.wanadoo.fr [193.251.34.75]) by mail.h3d.com (Postfix) with ESMTP id 6EAF3F9F1 for ; Mon, 22 Jul 2002 16:41:59 +0200 (CEST) Received: from h3d.com (unknown [192.168.201.21]) by smtp3.h3d.com (Postfix) with ESMTP id AE191B006D for ; Mon, 22 Jul 2002 16:40:42 +0200 (CEST) To: nfs@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Unsubscribe: , List-Archive: hi, i have 2 bugs with nfs. i recently installed 2 linux rh 7.3. one is 'the nfs test client'. one is mainly a samba server with acl support, all acl stuff (kernel, winbindd, nss, fileutils, e2fsprogs) is installed and seem work, windows users are happy :). On this server i have too configured a nfs server, the export directory is the same than root samba dir which has some special acl set, and there is no quota. bug 1 : on the client, after a successfuly nfs mount, when i try to access the mount dir, the process freeze and impossible to kill it. ex: ls /mnt -> freeze, kill -9 doesn't work, umount -f /mnt doesn't work. i used first the redhat kernel, i have tried with last 2.4.19 + linux-2.4.19-NFS_ALL.dif patch, update nfs-utils to 1.0.1, same result. don't think it will help, but i do some strace ls /mnt, many finish with : lstat64("/mnt", known bug ? a bad nfs server should not permently block a client, patch ? how can i kill these process without reboot ? is there something that i can do ? bug 2 : on the server, after a client successfuly does a mount, all 'basic' access (ls,cd but not samba) on the directory freeze the process as with the client, and the process is unkillable. kernel is 2.4.18 + acl patch. but i haven't try NFS_ALL patch or to update nfs-utils 1.0.1. strace ls /disk2 result : stat64("/disk2", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 open("/disk2", O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY) = 4 fstat64(4, {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 fcntl64(4, F_SETFD, FD_CLOEXEC) = 0 brk(0x8060000) = 0x8060000 getdents64(0x4, 0x805db20, 0x1000, 0x805db20 same questions, is there a patch ?? can i test something ? my goal is not to have nfs acl (i know there is nfs4), just want correctly mount and copy files. hope my english is not too bad. seb ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs